The enterprise automation landscape is changing, as issues such as the fragility of RPA and the proliferation of SaaS apps drive companies to look for new, more user-friendly forms of integration and automation.
RPA is one of the most talked about new technologies, but many businesses aren’t sure how to implement it. Those that have face high failure rates. In fact, according to Forbes, “an EY study found 30 to 50 percent of initial RPA projects fail,” and Deloitte reports that a survey of 400 global organizations found that “63 percent of surveyed organizations did not meet delivery deadlines for RPA projects.”
Automation has a historic trajectory, from the origins of the Jacquard loom in the Industrial Revolution, and the challenges accompanying automation have changed over time. Today, companies are striving to achieve digital transformation and business process automation, yet they often end up struggling, not only with implementing RPA, but with integrating the SaaS applications and solutions that they introduce to their tech stack.
The range of software solutions on the market is higher than ever, with 7040 for marketing alone, but are integration tools like RPA going to be able to keep pace with a rapidly changing business technology landscape? Industry-leading IT analysts like Jason Bloomberg have contemplated whether there will be universal integration of SaaS applications amongst the enterprise. In his new white paper From Fragility to Agility with Business-Driven Automation, he discusses the latest facets of integration technology, business technology paradigms, and considers what will be the future of enterprise automation.
In our Q&A, the internationally recognized thought leader on disruptive trends, SiliconANGLE and TechBeacon contributor, and founder of digital transformation advisory firm Intellyx, discusses challenges and trends he sees with both RPA and SaaS integration and why he thinks the latter will come out on top.
Read Bloomberg’s whitepaper, titled: “From Fragility to Agility with Business-Driven Automation” here >>
What are the biggest issues that companies are dealing with today regarding application integration and enterprise automation? Do you see common strengths and failures across organizations?
Perhaps the biggest issue is brittleness. Integration doesn’t take place in a context of stability. On the contrary, everything is changing — data formats, APIs, ephemeral microservices, and more. Too many integration and automation solutions simply break when so much is subject to change.
Can you give a specific example of an RPA system breaking down in a way that an Enterprise Automation Platform such as Workato would not?
Workato’s low-code approach empowers people rather than trying to replace them. So if there is a change that the automation can’t deal with, say a ‘breaking’ API change, then it’s simple and straightforward for a human to make the necessary adjustment.
Can you envision a future where there was a sort of standardized, universal system of integration programmed into disparately designed applications? For example, you can play any CD in any CD player, and fill up any gas-powered car at any gas station. Envisioning standardization for software.
Yes, in fact, I have hopes that innovation in the service mesh category will achieve this integration universality. Service meshes abstract integration endpoints, raising integration configurations to a declarative policy later. Today, service meshes are primarily used for east-west communication among microservices, but I envision broader applicability of this technology in the future.
You talk a bit about the inherent disconnect between potential citizen integrators and the business context necessary to compel integration. FW Taylor’s work arguably led to the rise of things like auto workers being expected to work like machines, without talking. Or essentially, mechanizing the body in ways that it’s physically and emotionally maladapted to. Complete digital automation, on the other hand, is very ergonomically viable. It is also an extension of Taylor-esque process engineering that breaks down business processes into a series of actions, governed by a script, and performed at top speed. You bring up his ideas to suggest that digital transformation realigns business processes with customer preferences, which “breaks down Taylor’s rigid distinction between management and labor, as leadership drives decision making responsibilities throughout a newly flattened organization.” I see here that the organization is realigning with the customers, but how does that flatten the internal organization of a company? Where do you envision this trajectory leading?
Taylor’s division between management and labor was a byproduct of the Industrial Revolution, but even to this day colors how people think about the nature of human work in large organizations. However, we aren’t in the Industrial Age anymore — we’re in the Digital Age. We must thus rethink this core hierarchical organizational principle in favor of bottom-up, horizontal organizations. This change is at the core of DevOps, and is an important enabler of Agile approaches. It is also at the heart of digital transformation.
Can you qualify the idea that the distinction from those who design processes and those who execute them is blurring? I would love to hear some examples of this.
Here’s an important way of thinking about this distinction. When you say ‘designing a process,’ you’re actually talking about two processes — the process being designed, and the process of designing a process. Traditionally, these processes are handled in different ways by different people, aligning with the distinction between management and labor. Today, however, the underlying process (say, a sequence of data entry tasks) is likely to be automated. So the only people we need are the ones dealing with the process design processes — and now, everybody is focusing on those and how best to leverage software to support or even automate them.
Do you think that empowering citizen integrators with a low-code or codeless platform cuts down on “shadow IT”? How so, and what are the benefits?
It depends. Certainly low-code/no-code tools can exacerbate the shadow IT problem — as did earlier generation tools like Lotus Notes and Microsoft Access. The difference today is that the better low-code/no-code tools have built-in governance capabilities that support IT’s priorities without slowing down the work of citizen integrators or citizen developers.
Can we talk about AI and machine learning? In its earlier days, computation used to be based off of symbols and rule-based logic. Machine learning makes it possible to analyze large, complex data sets and distill patterns from inputs, without being given the rules that govern the symbols. What’s your perspective on these issues?
We are beginning to see broader understanding of the different types of AI, including machine learning, natural language processing, deep learning, and a few others. As each of these technologies matures, I would expect clearer distinctions among them, as well as additions to the list.
Do you think that the move toward cloud-based subscription software and services is here to stay?
The subscription-based business model is now a core part of the cloud-native computing paradigm, so I expect it to be in the mix for years to come.
How do low-code RPA and intelligent RPA fit in to your discussion about automation? What fragilities do these types of RPA have in common with traditional RPA?
Much of what passes for RPA today is neither low-code nor intelligent. There is plenty of active innovation among low-code vendors to roll out low-code RPA solutions, but these are largely a work in progress. As for ‘intelligent’ or ‘cognitive’ RPA, there is progress with incorporating various facets of AI in the behavior of robots (say, to build virtual assistants or chatbots that leverage natural language processing), but we’re still quite a ways away from anything I’d call an ‘intelligent robot.’
What do you think is “next” in the enterprise automation world?
The current paradigm shift taking place in enterprise IT generally is the move to cloud-native computing. Automation is a part of this shift, and cloud-native approaches will replace older approaches to automation. Expect to see greater abstraction of underlying endpoints (see my comment on service meshes above), as well as broader implementation of microservices and a maturation of policy-based abstractions that drive automation behavior.
Get a free copy of Bloomberg’s whitepaper: “From Fragility to Agility with Business-Driven Automation” here >>
To learn more about Workato’s enterprise automation platform, request a demo from our team.
